Using Value-Added Information in Practice€¦ · Using Value-Added Information in Practice 5 making value-added decisions. And they had very little synergy occurring among all the

RECORD, Volume 24, No. 3*

New York Annual Meeting

October 18-21, 1998

Session 131PD

Using Value-Added Information in Practice

Track: Financial Reporting

Key Words: Financial Reporting

Moderator: DAVID M. BROWN

Panelists: ARMAND M. DE PALO

R. KARL ERHARDTt

Recorder: DAVID M. BROWN

Summary: Expert panelists look at how to maximize the usefulness of value-added

methodology. Session topics focus on how to leverage value-added information

and not on the actuarial issues of what value added is or how it is calculated.

Mr. David M. Brown: To date, most of the sessions have focused on what an

embedded value is and haven't really focused on how you use it internally in

helping manage the business. The idea of today's session is to focus a bit more

on using value-added techniques rather than what they are. Our first speaker is

Karl Erhardt from A.M. Best. Karl will give us his view, or Best's view, on how

value added should be used to help manage the company. Then I'm going to

speak about my experiences when I was in the U.K. using embedded values.

And, finally, Armand de Palo is going to speak about views of The Guardian,

which is currently implementing embedded values and value added, and why

the company is going through that process.

Mr. R. Karl Erhardt: I want to take a high-level approach toward value added,

describing what we're seeing in the industry today when we talk to companies

and then focusing on what value added means to me in practice, particularly in

terms of how the actuaries are influencing value added.

The companies we see manage for both short-term and long-term needs of the

organization. We'd like to take a long-term focus, as our ratings are focused on

claims-paying ability. This means the companies would be there 20-40 years

*Copyright © 1999, Society of Actuaries

tMr. Erhardt, not a member of the Society, is Senior Financial Analyst, Life-Health Department, at A.M. Best in

Oldwick, NJ.

2 RECORD, Volume 24

down the road to meet the obligations they are writing today. So we're looking

for the companies to be creating value in their operations and their daily activity.

That promotes financial strength, which is something that we base our ratings on

in terms of our models, which are static and similar to risk-based capital (RBC).

Obviously, those static models don't take into account the dynamic of real value

that would be derived from a company. So, in terms of looking at financial

strength, we're looking for the ability to sustain operating growth based on a

company's market profile and where it will be 5, 10, and 20 years down the

road.

We believe that value added is very important. It will tell us the difference

between the companies that will just survive in the marketplace and those that

will thrive, grow, and become the leaders of tomorrow.

Finally, value added allows us to compare the advantaged companies with those

that are disadvantaged. When you look at the companies today that are

advantaged in the marketplace, they look at value added as having a good risk

management program, having a strategy toward growth in the future, and being

able to realize cost efficiencies in their operations. In this regard, some

companies are advantaged over others. Companies with a larger capital base

and financial flexibility are much more able to move in and out of investment

opportunities and create value for the future.

When I was asked to speak, I put together something that would show we are

always talking value with the companies we meet with. And that came down to

four different areas. Policyholder value always comes up. Shareholder value is

probably the most talked about value-added topic in the industry today. And the

others are distribution and the blocking and tackling of your operation.

Policyholder value is derived from offering the highest possible coverage for the

lowest possible cost and knowing that the claim will be paid in the future. This

is where we're focusing in terms of our rating.

Shareholder value is interesting because we speak to a lot of consulting firms and

companies that say they are shareholder value-oriented. But what does that

actually mean? That means increasing stock price. We hear a lot about

economic value added (EVA) and market value adjusted (MVA), which is based

on the premise of earning above what your cost of capital is on a given

investment. We're also beginning to see management compensation, (other than

the standard stock options) tied into EVA. And some companies are looking at

market to book value.

3 Using Value-Added Information in Practice

In terms of distribution, we see a lot of value because growth in operations will

lead to a strong financial future. We ask companies all the time how will they

distribute their product, where are they distributing now, how much can they

sell through that, and where are their biggest profits going to come from their

distribution in the future? Companies that have built a distribution system are

more advantaged than those that are struggling with their current high-cost

distribution channels.

Finally, in terms of operations, we look at productivity and efficiency in the

everyday operation. Sometimes this is overlooked in terms of value, where

value is mainly focusing on senior management.

Let's step back to shareholder value, since it is the most discussed topic out

there, and consider how wealth is created for the shareholder. There are some

inherent contradictions in this because it's meeting an analyst's expectations, and

analysts expect stable earnings growth and projected growth. So, the question

then becomes, are we managing for the short-term three months out or making

our decisions for the long term?

I'd like to focus on the difference between short-term and long-term

management. When a company is truly focusing on value, it is looking toward

the longer term, as opposed to meeting the short-term expectation of the analyst.

"What am I going to do for the next quarter to meet my expectations? I can do

acquisitions and the like." This has happened in the past. And recently, we've

seen companies struggle and fall to the ground because of managing to

expectations in the short term.

So, in light of that, we like to see companies managing for economic value-

added. Economic value-added, to me, means real value added. When we talk

about MVA and EVA, we tend to come to the market focused on the long term.

And I try to think of it in terms of real value-added, meaning: What is my cost of

my investment? What am I including in a risk piece? And what am I earning

above that? This is opposed to going to the external marketplace to measure my

performance.

Think of the tortoise and the hare. Obviously, the tortoise is the long-term

company. A company called AIG has purchased Sun America. There are a lot

of interesting things happening with that transaction in terms of how it is being

recorded. For one, it's going to be a pooling of interest, which is an interesting

value-added piece. Within that transaction, there is real economic goodwill

occurring. In terms of the value of that transaction, you're at $18 billion. And

the company is now worth $4 billion, so there's $14 billion in terms. Although

4 RECORD, Volume 24

there's an accounting difference between purchasing and pooling, some

goodwill or real value was also acquired. They say the value they acquired is

worldwide distribution. And they think that they will be able to make that go

forward and realize all that value. Why that's important to me as an analyst

thinking about real value is that, if I amortize that over a 30-year period, I

actually wipe out more than the company is currently earning. So, obviously,

they're projecting large growth going forward. I would certainly not question

whether that's possible.

On the other hand, there are companies that go out and buy blocks of business

and say, "We're going to earn a good ROE on this." They pay a lot of money for

a block of business that ends up costing them a lot down the road. To me, that

decision-making process did not take into account the real value or the risk

associated with that business.

So, to me, the difference between the tortoise and the hare is actually a company

considering its reasonable risk-adjusted returns versus what the market expects it

to do. In acquisitions, the tortoise would be the company that goes out and

makes a good purchase and the hare would be the company that makes a

purchase to increase earnings in the short term.

We see value added occurring in two places. We see senior managers talking

about it in annual reports, 10Ks, and the like, because that's part of their job.

Part of their job is to provide strategic vision to the company and put together a

business plan that moves the company forward. Their compensation is going to

be based on their success. When we look at compensation that really is tied to

performance of the company, it generally comes in two parts: stock options and

deferred compensation. And usually that's tied into the market return, which

inherently provides a little bit of short-term/long-term management indecision.

There are companies that are looking at a longer-term value-added approach to

compensation, but it's not widely practiced.

The part that gets overlooked most when we talk value-added to our companies

is coming up with performance measures and drivers of business functions and

processes that can be improved and actually add value to the organization. My

biggest example of this is a company whose stock price dropped recently from

$40 to about $1.50. It was a major acquirer, and senior management actually

had been replaced at this point. The managers talked the short-term talk. They

were able to accrue earnings through acquisitions and hide their underlying

operations. Yet, they never focused on improving their functions, such as getting

the best technology to produce strong management reporting data essential to


making value-added decisions. And they had very little synergy occurring

among all the companies they added.

Value added starts with reliable and accurate information that can enable strong

decision making. That comes in two parts. It comes from management

performance information and a strong risk-management program. The

companies that we see adding value in their decisions can show that to us very

readily because they have the information available. Many companies do an

acquisition and yet have no information on the company. So, in terms of

looking at management performance information, companies are going through

their businesses and coming up with key performance drivers that they can

measure that business with and then allocating capital based on performance.

The information should be timely and accurate, not six months after the fact, the

company should be proactive instead of reactive.

The other piece is risk management. I would say those companies that we see

with the best financial reporting are in the process of developing a good financial

risk-management program. What I mean by financial risk is going through your

interest-rate risk, market risk, and credit risk from an overall enterprise basis and

then looking at what I would call insurance risk for the products that they're

involved in. They also look at the drivers of where they would like their

business to be in terms of a spread of risk and developing a risk appetite. The

companies doing this are going through the process of identifying what their

risks are, defining those risks, taking action to manage to those risks, and then

following through with financial reporting.

Other companies are adding, on top of that, identification of their business risks,

which is a hard topic to discuss because it varies from company to company.

But it's looking at running the daily operations and how the company expected

results will occur.

The next thing I'd like to touch on then is the actuarial involvement in the

organization and where it's adding value. The main areas are: the pricing,

which is forward-looking; the productivity of the company, everyday operations

meeting and exceeding the assumptions of the company; being involved in the

capital management process; and then focusing on the holding company.

In terms of pricing, it starts with better information. And I think my counterparts

are going to address that. But, what I wanted to talk about is the difference

between statutory, GAAP, and economic reporting. I want to identify what we

most commonly look at and have to analyze in terms of real value. The main

difference is in the methods of accounting that we end up focusing on: the

6 RECORD, Volume 24

realized ability and accuracy or treatment of any deferred acquisition cost (DAC)

or value of insurance acquired.

With respect to pooling versus purchase accounting, it's very difficult to look at

companies on an even keel because they are accounting for their major

acquisitions in different ways. I can look at two different companies with two

totally separate sets of balance sheets. They may actually be identical

companies, but look so much different. From our standpoint, how do we look at

the company that's truly adding value? When do we start to discount that

amount of goodwill, that amount of DAC that was acquired in an acquisition,

compared with the company that doesn't have it on its books? Obviously, value-

added reporting can help that.

We look at income from continuing operations versus a total return. How much

credit do we actually give for real life gains within an operation? Obviously, real

life gains are part of our pricing, but not entirely, especially in spread businesses.

So how do we normalize that amongst companies we look at?

The final thing we examine is FASB 115 adjustments. They're the major pieces

that show a difference between statutory and GAAP and the true value added in

a company.

Companies doing the best pricing are using the best technology available. I'm

not here to advertise for those companies, but it always comes back to several

different actuarial systems that companies are using. From my standpoint, these

systems give them better information for their reporting process. I look at a lot of

actuarial memoranda from companies, and it's very easy to see companies that

are using the latest and greatest software to determine their liabilities. They are

actually strengthening reserves. Then we put that next to companies whose

actuarial memoranda look 10 years old. They're talking about, "We might need

to strengthen," or what have you, but they really don't have the information

available to know the exact liability. To me, there is value added to the external

community in knowing what that liability is, booking it, and moving forward.

In terms of meeting and exceeding productivity assumptions, the companies that

are doing the best job of coordinated reporting are those with the best decision

making. They are combining their actuarial needs with the management

reporting and coordinating that all within the operations and investment

management areas. That's a big circle of knowledge. Everyone knows what is

necessary and management then can make its best decisions.


That's important because value added is a better method for making decisions

than, say, ROE at times. ROE and return on assets are two of the most

commonly used methods. But in determining my value added, I have to come

up with my cost of capital. My cost of capital for the company is, say, 11%.

Most of the companies in the public domain are probably between 10-12%

under the current economic environment. A lot of companies are only earning

7-8% on their life businesses. If they can write more business at 9% on a ROE

basis, they're going to do that. On a value-added basis, however, their cost of

capital is 12%. Therefore, they're not adding value to the enterprise. They are

actually taking away value in the long run. In contrast, they could be earning

20% on a business and not invest in something that did 15% because they do

not want to reduce their ROE. Yet, it's something that would add value to the

company.

In terms of capital management, I like companies that are allocating their capital

based on their liabilities and surplus. They look at strong asset/liability modeling

(ALM) against the insurance liability and then identify what the required surplus

is to manage those liabilities. Then they identify excess surplus. What you tend

to see then are investment policies that say: assets to match are limited to fixed

income, "A" or better investments; then they take on a little more risk with the

required surplus-maybe BBB,. a small junk portfolio and some mortgage loans;

and in the end, in the excess surplus they take on higher risk investments.

The only caution in this is the difference between insolvency capital and

economic capital-the capital required to move this business going forward in a

stable earnings stream. So this is probably one of the best ways that we see for

companies to combine their capital management program, and, through that, to

allocate their capital effectively.

It's interesting when we look at ALM. Many companies tell us they do it only for

the regulatory reporting requirements. Then we talk to a little better company,

where managers say they coordinate ALM with their investment management.

Better yet are those who integrate it with their financial risk management and

those where it's part of their decision-making process. In some it's part of their

strategic planning process, where the actuarial department is highly involved in

the long-term strategic planning of the company. Obviously, the best practice in

the industry includes all of the above. But there are companies that focus at all

different levels on that. It's our opinion that ALM is a very important part of

capital allocation.

From the high-level standpoint, understanding the business mix of the company

is also important to adding value to the company. I'd hate to say that companies

8 RECORD, Volume 24

cannot invest in their life insurance operations because they produce a lower

ROE or create less value. We have to understand the puzzle that the company

is, put it all together, and understand why we're investing all of that. So maybe

the best value added to the company is actually not investing in what is the best

value-added investment option based on the overall enterprise. Then, it's

considering the enterprise risk and what is in there. And that really comes into

our cost of capital calculations in terms of analyzing or giving the risk

component of capital.

Finally, we think that value added is a long-term process. Obviously, you have

to balance your short-term and long-term objectives. But history shows us that

focusing on the short term limits our options going forward and in the end,

probably destroys value. So short-term companies tend to focus just on short-

term growth and increasing market value, not on economic value. They tend to

do deals to please the investor community, e.g., cut out $15 million or $20

million of expenses in December to add two cents per share earnings.

Companies that focus on the long term build their fundamentals and prove their

operations have very strong internal pricing and compensation systems. They

have a focus also on the policyholder. In the end, we believe that highly rated

companies will have a strong value-added incentive within their organizations.

With that, I'd like to turn it over to Dave.

Mr. Brown: I'd like to talk about what I see as the main strength of value added

and then show how that can be used to help set and validate the strategic

direction of the company. It can then be used to help manage performance on a

day-to-day basis. By integrating those two-the strategic direction of the

company and the day-to-day management-you can ensure that your

management efforts are being aligned with where you want to go in the long

term.

Let me just quickly identify what an embedded value is, because it's the main

value-added tool that tends to be used in life insurance. An embedded value is:

the present value of distributable profits from the in-force business; plus adjusted

free assets. The value of the in-force business risk is allowed for in that

calculation by discounting at the risk-adjusted discount rates. And the

projections are done on best-estimate assumptions.

There's nothing magical about this methodology. It's the same methodology

used in pricing-that any business school person would use in a capital project

appraisal. It's a tried and trusted methodology. What's interesting about

embedded values is that, when you look at the change in embedded value from


year to year, you can break it down into four components. There's the

unwinding of the risk discount rate or the cost of capital on the opening

embedded value. There is the value of new business written in the period. And

then there's a piece of variance on the in-force business, which management

inherits from the prior period. If that variance is a large number, it's indicating

weakness in the model.

Because the same profits rise on all three measures-statutory, GAAP, and

embedded values-the important thing to understand is that, when the last

policy goes off the books, the accumulated earnings are the same under all three

methods. The difference between the three methods comes in when you

recognize the profits. Under embedded values and value added, when you

write your business, you immediately recognize the full economic worth of that

new business. At the other end of the spectrum is statutory reporting where it's

30 or 40 years before you effectively recognize that writing new business was an

economically sensible thing to do. And then you have GAAP somewhere in

between. If you're using GAAP or statutory, then that means you're not

recognizing the full consequences of writing new business in the period today.

This means that the performance of future periods is going to be influenced by

the actions you're taking today. Another way of looking at it is that your

performance today is essentially being impacted by events and actions that were

taken over the last 5, 10, and 15 years.

If you then consider the environment we're in, which is rapidly changing, and

much harder to price on a cost-plus basis, it's more market-driven. We're

making bigger infrastructure investments. And it's much easier to make

investments if you are given an adequate return. So if we use an accounting

process that tells us how we were doing 5, 10, or 15 years ago, it's totally

inadequate. The strength of value added is the fact that it tells us fully the

consequences of actions and events that are happening today.

Let me give you a quick example of how value added might differ from GAAP.

The two main pieces of information that managers are interested in is the value

added on new business and the value added on in-force business. So, you could

provide a high-level summary by the different lines of business of the

performance over a period of time. If you just focus on annuities, for one

company, the annuity line of business destroyed value to the order of $35

million; $33 million was lost from the value of in force and $2 million was lost

on the value of new business written in the period.

The company that we were doing this work with had previously been looking at

this business on a GAAP basis. On the GAAP basis, there was GAAP equity of

10 RECORD, Volume 24

$500 million. Company managers had a benchmark of 13.5%, so they wanted

to achieve profits of $67.5 million. They had operating profits of $68 million,

and as far as this company was concerned, on a GAAP basis, everything was fine

with the annuity line of business.

When we looked at this on an embedded value basis, there had been an

increase of value of only $38 million. The benchmark was $73 million, so there

was value destroyed of $35 million. The management of this company had a

completely different perspective on this line of business.

Here are the reasons behind these two different results. The annuity line of

business was becoming more complex and the company had to invest to service

the business. GAAP was only bringing into account the extra maintenance

expenses in the current period, whereas value added is saying, "These extra

expenses are going to occur in all future periods. Are you taking into account

that much higher cost?"

The company was having problems with sales halfway through the year, so it

introduced a higher commission rate to encourage sales. Under GAAP, this

higher commission is spread over all future years and has very little impact on

the current period. With value added, you're taking into account the full impact

of that commission now to show when new business is impossible.

And then surrenders were increasing. Surrenders were giving surrender profits in

the current period, which improved the GAAP results. Embedded value takes

the current improved statutory earnings, but offsets this against the lost margins

in the future to show that the surrenders are an economically bad occurrence.

Once you have an embedded value model up and running, there are two ways

of using it. One is that you set a value at a particular point in time and run the

model on the different scenarios. Management is thinking about course of action

A or course of action B. You can then use the embedded value model to come

up with a value on those two different scenarios and identify which is the better

option to take.

The other way of using embedded values is to look at performance over a period

of time. You have a value at the start and end of the period. You then analyze

what caused that change in performance, which helps management understand

the external and internal events that are impacting the company. It then allows

managers to react and better manage their company.


In the U.K., there were lots of new companies sprouting up about three or four

years ago. So some of the banks were setting up a lot of subsidiaries. The banks

and the subsidiaries didn't want to show a series of losses and wanted to have an

accounting system that fairly reflected where they were. They decided to use

value added. The managers wanted a process and a way of looking at a the

business that focused all their thoughts around managing to create value. The

basic idea is that senior managers think in terms of the long term about where

they want to be and in which direction they want the company to go. They then

think of the capabilities and the characteristics of the company that will succeed

in that future environment. Then they develop a plan to change the organization

from where it is today to where it needs to be in the future. Finally, they need a

way of monitoring whether the plans implemented are successful.

You monitor that through key performance indicators. The idea behind key

performance indicators is, if you're managing to create value, then ultimately

some of your economic value drivers must change. If your economic value

drivers were things like lapses, expenses, and mortality, to create value,

ultimately they must change. However, if you think of mortality and lapses,

sometimes it's two, three, or four years before your experience comes through

and you can identify whether the management actions have actually had the

intended benefit. Key performance indicators are a softer way to identify that

our plans are having the success we want without having to wait years to

measure the hard financial statistics.

Value creation should be underlying the way management thinks in terms of

developing the strategy. You should then be running the models to validate that

the long-term strategic direction is correct and viable. You should be using value

added at the ends to measure and check that the value you intended to create is

actually being created. If it's not, it allows you to take corrective action, thereby

using embedded values with a strategic direction.

When you start running scenarios, forcing somebody to think through all the

assumptions to ensure that everything makes sense, adds a much higher level of

discipline to the whole process. When senior managers have discussions,

having hard numbers in front of them about the benefits and the costs involved

leads to much better decision making. It moves the discussion from being

idealistic and subjective to being much more objective. And if people object to

the proposed scenario, because they have the assumptions and beliefs laid down

in front of them, it's much easier for them to be informed and to criticize or

agree with the proposed strategy. The numbers that have been used are much

more rigorously calculated. Many people will use simple spreadsheets or back-

of-the-envelope calculations. Clearly, these can be wrong or incomplete. Doing


the calculations creates an effectively laid out plan for how the benefits are going

to arise. This allows managers to have much better accountability and the ability

to recognize when things are going wrong, which allows them to take corrective

action.

This is how value added can be used to help set the strategic direction of the

company. It also helps managers understand their current performance.

Understanding current performance allows management to react to it and to

make decisions to correct poor performance. So the classical way of using an

embedded value is, each period, you monitor your current experience in terms

of lapses, expenses, and mortality. Through monitoring the experience, you then

update your assumptions to get your best-estimate view of the future. Then, at

the end of the period, recalculate the value of the company and analyze the

reasons for the change in the value of the company. Some of you have actuaries

pricing the products who are claiming that they are pricing the products to

achieve 12-15% rates of return, or whatever. If they're achieving that, then it's

validated by the process. If it's not, it forces them to come out with a business

reason to explain to senior management why they're not getting what they said

they were aiming to achieve.

Although some people view embedded values as a tool that's just for actuaries, I

would argue that it more brings actuaries out into the business arena. Because

value-added information, as we saw before when we were looking at the value

added by different lines of business, is very simple. If it's a negative number, it

forces the business people to ask questions about why profitable new business

isn't being written. It's better than finding out that we're pricing to achieve new

business profits of 12% using expense allowances but, unfortunately, the costs in

our company have changed and the expense allowances are no longer correct.

Three or four years ago they were correct and we just haven't bothered to

change them. That sort of thinking is brought out into the open. And senior

management has a much better understanding of the practices that are going on

in the company.

Similarly, you should be integrating the planning process within this procedure.

If management is spending lots of time taking actions to try and improve lapses,

the fact that no value has been added by the change in lapses over the year is

equally important. It's good for managers to realize that all their effort is leading

to no improvement. This can then allow corrective action.

So the way to provide information that management can use is to go back to the

definition of the change in embedded value I used at the start. There are four

pieces: the unwinding of the discount rate, the value of new business, the value


added on the in-force business, and the variance on the in-force business. The

variance on the in-force business consists of two pieces. One is modeling error.

Life insurance is a complex business and the models we build can't always fully

capture the dynamics of the real world. The other is where you may have a

long-term assumption for mortality that's a good assumption, but year after year,

clearly, you're going to have fluctuations in your mortality experience. If your

mortality is lower than your long-term assumption, that leads to higher profits in

the year, which boosts your free assets and leads to an increase in value. What

tends to happen is you have a deadline for providing the embedded value

numbers, so a team of people works hard to get the numbers out. You

mechanically work out what the value of new business is. That's easy. You

work out the value added on the in-force business. That's reasonably easy

because you had an ultimate lapse of 10% a year ago and now have an ultimate

lapse of 9% for this year. It's pretty easy to work out the value added by that

change in the ultimate lapse rate.

The hard part for the actuaries who are doing the work is understanding the

variance on the in-force business. You're trying to identify any weaknesses in

your model and correct them before you present the numbers to senior

management. Unfortunately, you've just come through this process and spent all

your time focusing on the variance on the in force. What often happens is that

the information you present is what you've been spending the most time on,

which is the variance on the in force. But that's of the least relevance to senior

management. What you need to do is focus on the value of new business, the

value added on the in force, and explain to management what external and

internal forces are acting to either create or destroy value. You need to think of

how is real value created. Things happen in the real world. Competitors change

the way that they're competing. The economic environment changes. New

regulations are introduced that lead to extra expenses within the company.

Consumers change the way they behave. And in company management, you

implement various plans that are going to change the way the company behaves.

So the external and internal actions and environments change the operations.

For example, if competitors are increasing their commission rates, that can

impact your agent retention. Or if consumers prefer investment products, you

start seeing a shift away from life sales to mutual funds and annuities. This leads

to new volumes and expenses, etc., which leads to economic value being

created or destroyed.

Then your analysis goes back the other way. You do your analysis at the end of

the year, which tells you how much economic value is being created or

destroyed. You then identify where that's being created-in new business,

expenses, mortality, or surrenders. Typically, the various analyses that I've seen


tend to stop there. They analyze value being created in terms of either expenses,

mortality, or surrenders. And that's as far as they go. I think you need to go

back further to identify the operational drivers that have changed. Keep on

breaking down the sources of profit or loss until you get to a level management

can react to. You need to get to a level where you can make somebody within

the organization accountable for the underperformance or the overperformance.

You then need to identify the external influences that are impacting the

company, which then allows management to react to those drivers.

An example of the difference can be seen when you present value being created.

Your variance analysis may say you've lost money because of your in-force

product structure. You've reduced your spread on your in-force business.

Expenses are higher, so you've lost value on higher expenses. But surrenders

have improved, so you've made money because of the improved surrenders.

That would be the way that a typical variance analysis would be presented.

The opposite way is to say, "Our competitors have all been coming out with

much more competitive credited rates. We have no choice but to reduce our

spread to stay in the market. We're trying to differentiate ourselves by having

extra service and introducing technology to differentiate our service to our

customers, which has clearly led to higher expenses. But then that's been offset

by improved lapses, so our analysis of the market and our actions were

validated." Instead of looking at each discrete piece, we've brought them all

together and given management a much better feel for what's actually

happening. And that's where I'm going to end.

Mr. Armand M. de Palo: I'm really the practical person, because I'm the one

who is actually in the process of implementing an EVA system in The Guardian.

I want to start out by saying Guardian is a mutual company. This is not a

method just for stock companies. In fact, being a mutual company, there may

actually be more of a reason to do EVA than for a stock company because I don't

have a stock price outside the company to validate whether the company is

increasing or whether somebody perceives the company increasing in value.

The Guardian has looked into this method and we're doing it for two reasons.

One reason is that, in today's world, we want to link part of senior

management's compensation to the increasing value of the company, so we

needed a viable way to measure it. We looked at the accounting systems that

exist and found that they don't do a very good job. The other answer is to get

closure on the pricing of products so that managers get information from the

actuaries that lets them know what assumptions are being realized and if there is

total integrity to the pricing. You're going to find out that the gain and loss

analysis is probably one of the most valuable tools in this whole thing. It's


overlooked by most people because they don't appreciate that the analysis tells

you whether you are achieving your pricing assumptions or not. And it also

gives you information as to when you should be changing your pricing

assumptions.

First, we had to sell management on the idea. It's not cheap to implement EVA.

It's another accounting system. So the first process we went through was having

to address the following question: "If you already have statutory, tax, and GAAP,

why do you want another accounting system?" We spent a lot of time on that

subject and we concluded that none of the existing systems gave us information

about whether we were increasing the real value of the company.

Here are a couple of quick examples. Statutory reporting uses conservative

assumptions. It really doesn't give you any information on the economic value

of the company. The simple way to view it is that you write new business; new

business looks like a large loss to the company. Maybe, for some companies,

new business is a loss. But if the business is profitable, it's adding to the

economic value of the company. And the reason you know that is that you

could turn around and sell a block of business to the reinsurers and make a little

bit of money off of it.

Tax, which is another accounting system that runs parallel to statutory, still works

on relatively conservative assumptions, even though some aspects of it are less

conservative than statutory. It has a deferred acquisition cost (DAC), which has

nothing to do with the real value of the product. It's just in there to generate tax

revenue to the government. Once again, new business looks like a loss on this

system. And it still doesn't tell you anything about the real economic value of

the company.

Some people are misinformed about GAAP; they think it is telling you something

about the value of the company. While it uses more realistic assumptions,

GAAP is trying to get at a different answer than the economic value of a

company. It's trying to determine consistency between companies and is driven

by accounting methods that are set by nonactuaries independent of the pricing of

a product. It makes no attempt to be consistent with pricing. The one advantage

you have with economic value is that you can set the assumptions so that they

are consistent with how you're pricing your products and with the assumptions

that go into those products. That's a very informative feature because GAAP

doesn't do that for you.

When we were looking at tying compensation into it, one of the things that came

up was, if we use GAAP, how will we use it? And the conclusion we had from


consulting was that if you used GAAP, you'd have to go to the real marketplace

and find out what multiple of GAAP you'd assign to economic value. If you've

ever looked at stock values, they are normally some percentage of GAAP equity:

120% or 150%. But it really became obvious that, if you were grossing up

GAAP equity like that, you were trying to get to a realistic value because GAAP

wasn't doing it.

Value added does this for you. With this method, we can set the hurdle rates

that are consistent with the cost of capital. I will admit that, in a mutual

company, cost of capital is a different animal than in a stock company. A lot of

your capital is internally generated. And you're probably going to have a lower

rate than if you had to go outside for capital. However, if you have to go outside

for capital, you cannot on those blocks of business continue to say you have

internally generated capital. So you have to make sure you're consistent on that.

Also, in a mutual company, there's a lot more thought given to after-tax earnings

and accumulation of surplus. Economic value is a consistent method which has

statutory, after-tax, and accumulation of surplus. It's just discounting it. So in

many respects, it's a very natural method for a mutual company to assume.

We've been working on this for over 12 months. We've had little models

running, but we're planning to have a full implementation of the system by

December 31, 1998.

Many people view it as a tool of the actuaries. That's going to be your largest

problem. When you're first out of the box, the question is, "Why does the

actuary want this?" The accountants are not going to give you a lot of support

because it's not under their control. And the accountants will try telling

management, "We have GAAP already. Why do you want to spend an extra

amount of money to do all this? We don't support it because the FASB tells you

that these are the standards, so why do you want to go in this other direction?"

You have to be communicative with management. If you don't show the merits

of the method, you're not going to get the funding. And the funding is fairly

substantial on this. You have to have an actuary working on each product line

for about 60 days a year once you establish the method. And they really can't

find the time to work on it until they close out the annual statements.

We're going to do it once a year after the close of the financial statutory and tax

statement. We're going to produce the economic value pretty much in the

March time frame. But the gain and loss analysis is a longer-term process.

There's a summer deliverable on it. You've got to tell management all this up


front. If you say, "Oh, it's another tool. It's not going to cost you very much,"

and the bills start coming in, they're going to close you down.

It's also important to recognize that you have to reconcile with statutory. That's

a very important feature of this, that there's reality tied into this model. You can

start out with an assumption, but once a year you're going back to a real in force.

You're going back to a real amount of cash value. And you're showing the gains

and losses that are going on. It's important to realize that if you only are looking

at the yearly swing, you don't want this method. This method swings much

more on a year-to-year basis than any other method you'll look into.

You're really trying to communicate to managers the long-term trend of the

growth and economic value. If you say, "One year the value went up $100

million, and the next year it went down $80 million," they might not see the fact

that what you really have is a change of $20 million a year. On a trend basis,

that gyration could cause them problems that they don't understand. The reason

that the values gyrate is experience doesn't come in smoothly each year. Some

years your lapses are going to be better than others. Some years your mortality is

going to be better than others. And all those are going to run through your

model as yearly gains and losses. You have to look at the yearly gains and losses

and decide when a variation becomes a trend. When a variation becomes a

trend, you have to change an assumption. That causes an even bigger swing

because, once you change an assumption, the whole present value of that stream

is affected by that change in assumption. You have to explain this to

management up front.

What does it cost to do a system? Guardian has lots of tools already in place.

We own most of the software that's in the marketplace to value products

because we're in the assumption reinsurance business. So we basically own any

system that most other companies are using. We can run PTS, TAS, and

PriceWaterhouseCoopers' software. I'm not sure a lot of other companies have

as many packages. This is unique to The Guardian because of our reinsurance

operation.

If you don't have existing software packages and have to build from scratch, it's

going to be more costly. The up-front cost on this was several hundred thousand

dollars. And it probably costs a couple of hundred thousand dollars a year to do

the work each year. It's not free. If you're going to use it like we plan to use it

for compensation, you probably want your external auditors to review the

method each year for consistency. It's not GAAP and they won't give you an

opinion that this is consistent with GAAP. But you do want them to do the

analysis showing what you did one year is the same as you did another year, and


that any changes you made in either method or assumptions were adequately

reported to management.

You then get into the gain or loss. I'm going to go back to the gain and loss

because some things need to be broken out separately. If you use the model to

create a compensation system, like we are, you can change your model. One

year, you may have a small model. Guardian's experience with modeling is that

you start out with a small model, a couple of thousand policies, and the models

tend to grow on you. As they grow, they get more exact, so you haven't really

changed the economic value of the company because you've come out with a

better model. Whenever you change your model, though, you have to run it on

the old system and the new system, and calculate the Delta. And since we're

keeping track of units, you have to increase or decrease the units any time you

change the model so that the value per unit doesn't change.

This was a very important decision for The Guardian, because if you're going to

tie compensation to economic value and someone changes the model so value

goes up or down, you'd have a problem. In essence, compensation would

change even though value, in reality, hadn't changed.

You also have to make a decision on whether a change in future assumptions is

something that you want to break out separately. We decided that it would drive

the value of the company, so we want to communicate that in a separate item

from gain and loss. If you change your mortality assumption, the whole present

value comes through, or the change in lapse assumption all comes through. But

you have to make that very clear to management. And that's when you want to

link it to the yearly gain and loss. If you're seeing a trend, you have to have an

ongoing discussion about when the trend requires a change in method.

When calculating the gains and losses, we decided that we did not want to bring

new sales into our EVA method. This is probably not a decision that a stock

company would make. Our EVA method is an in-force method, and new

business comes in as a separate item. We will report a gain and loss in value

each year that is a result of new sales. If you included sales in your EVA model

directly, you would have a future sales assumption that went out into the

hereafter that said "This company is going to sell this amount of business in each

future year" and you would value it. And if you sold more or less business than

your model said you would in the future, your economic value would be

increased or decreased because of a variance from your expected sales

assumption.


If you were going to sell a company, you would probably put the sales

assumption in. We rejected this because, once again, we were going to tie it

into a compensation system. It was too tempting for someone to say, "The EVA

is not high enough. We're good salespeople. Let's crank up that sales

assumption a little bit more so the EVA will go up." We felt that it was much

more reasonable to let the management see the gains and losses on new sales.

We also think that was an important feature. If salespeople write new business,

they are going to see the present value of profits on that new business because

that's not in the EVA. So you have to value the gain and loss in the calendar

year itself. It will tell them they wrote a piece of business that's going to produce

a profit for them. This was important to The Guardian because we have multiple

lines of business that are quite different from those of many other mutual

companies. We are in the reinsurance business. It's not widely known that The

Guardian is in the reinsurance business because we don't sell facultative and we

don't sell automatic. It's basically block reinsurance that we do for other

companies. But we have about 60 reinsurance treaties out there, and the

question always comes up: "Are you making any money on these treaties?" The

EVA system is an ideal method for us to keep track of each of the reinsurance

blocks to see if it repays the investment we made.

The reason I bring that up is that we have different hurdle rates for different

products because we're risk adjusting the business. We're starting with the same

basic earned interest-rate. Guardian right now has an internally generated

capital source, so we're not looking at outside indexes for capital. We're using

an internally generated cost of capital. But participating business is basically

where we start. Guardian will do some reinsurance for window dressing of our

own surplus levels. These are very, very low risk corporate decisions to manage

the amount of surplus that we have. We use a slightly lower earnings-on-surplus

type of ROI to ensure that these agreements repay the surplus they entered into.

In the par business, we assume there's a small amount of risk, but the dividend

itself lowers the risk aspect. And the non-par lines we have are mostly variable

life and variable annuity lines, so we use a higher rate. We've agreed on what

the spreads should be. So if we change one rate, the spreads stay constant and

we just scale them all up. But we do have three different hurdle rates we're

using in the model based on the risk profiles of the business.

We also needed to look at the question of including target surplus or not. The

first guess on that was to just go with some multiple of RBC as the target surplus.

But we had multiple lines, and there's a square root function in the RBC

formulas. How did you set it so that it's stable by line? So we decided to do


something else. In free surplus, you want to keep track of the statutory minimum

amount of RBC and, depending on your convention, that's either the 125% or

the 250% level that you don't want to fall below. We're assigning surplus by

line of business on a fixed formula method that's less than 250% of the RBC.

We think that's the more realistic surplus for our company. And the excess not

assigned to the profit centers is subtracted from the free surplus of the company.

When you look at all EVA, you have the economic value of each existing block.

You also have retained surplus. One of the questions that comes up is how

much retained surplus is really available for new investments. And you really

don't have the amount that has to be held for economic purposes and for RBC

purposes available to invest in another line unless you go to outside financing.

We also are a dividend-paying company, where capital gains enter into the

calculation of dividends. If there are capital gains, you would see an increase in

value of the company on both a statutory basis and a GAAP basis, especially in

common stock. But if it's your practice to pay your capital gains out to

policyholders, whatever percentage of the assets belong to the par lines of

business, those capital gains will be paid out in the future. Therefore, they don't

increase the value using an EVA method. In effect, a large share of the capital

gains that we have don't increase the economic value, but increase the future

dividend that will be paid, and that's recognized in the EVA model.

On variable products, you have to make a decision about what the underlying

funds are going to earn. We decided that the answer can't be derived from what

we've seen in the last year or two. We have to look at what we believe will be

the 20-year return on those types of assets and report the yearly variation when

the stock market either goes up or down by a large amount.

Even though the method is not fully implemented, each line has run a small

model. And all the actuaries are now building a bigger model. The question is,

"What size model should they implement?" I believe it will take three years

before we are completely finished with the implementation. While we'll be

operational, we'll still be tweaking the model for probably the next three years.

Some of the early information that came out of it would probably have occurred

to people if they had thought about it, but the model made it clear. We have

two products in the equity line, a variable annuity and a mutual fund. If you talk

to the marketing people, a dollar of sales in variable annuities looks very much

to them like a dollar sale of a mutual fund. But when you do the EVA on it, you

get a very different answer. The mutual fund has much lower profit margins and

the funds are less sticky and have much higher turnover. The funds in a variable


annuity have a mortality and expense charge and some underlying fund charges.

So the effect of after-expense profit on the variable annuity is substantially higher

and the lapse rate on the variable annuity is substantially lower.

We found from real experience that a dollar of variable annuity has four times

the value of a dollar of mutual funds. This tells that, in order to get the same

critical mass in a variable annuity line of business, you only need one-fourth the

assets you would need in a mutual fund business. So if you start with the

assumption that you can't be viable in mutual funds with less than $10 billion of

assets, you may be viable as a company with $3 billion worth of variable

annuities. This wouldn't be immediately obvious to people because the

salespeople will say they brought in the same amount of money. This was an

important issue.

There's other early information that we're getting. The board has gotten the

message. Managers are very interested in the gain and loss analysis because it

reveals whether you really achieved your assumptions. Did managers spend the

budgeted amount of expenses they said or go over? I have a hunch this model is

going to give us a real good handle on pricing for expenses. And, even though

we have asset shares and the actuaries know we're covering all expenses, there's

no good tool to communicate it to management. This really will be a solid tool

to do it.

Even though from an actuarial point of view, this is the tool I've always needed

for gain and loss analysis and information, if it was not tied into the need to do

something on compensation, I probably couldn't have gotten enough support

from the rest of management and the actuaries in each of the different profit

centers to get this model in.

Mr. Isadore Jermyn: I have a number of questions for the panel as a whole, but

I'm particularly interested in Armand's response to them. First, Armand, you

mentioned that sales were excluded. It wasn't totally clear to me. I assume you

mean future year's sales. I assume the current year's sales are included.

Mr. de Palo: That's correct. There are two ways to do EVA. The quantity we're

using is EVA at each point in time. So if you're in year T and move to year T+1,

obviously you have one more year's worth of sales. So the EVA at the next year

has the sales. What's coming out of the model is more useful than if you

included sales. If you included future sales in your model in EVA, you'd be

reporting the gain and loss because of the variance in new sales. But what we're

reporting on is the gain and loss: Did the new sales add to or subtract from the


value of the company? And, for our structure, that's an important piece of

information.

Mr. Jermyn: My next question is in regard to the policyholder dividends.

Decisions can be, and sometimes are, made about changing policyholder

dividends independent of obvious changes in experience. How would that be

factored into the determination of the value, particularly as it affects incentive

compensation?

Mr. de Palo: If you lowered the profit margins, it would lower the EVA of the

product, and the EV of the company would come down. One actuary in my

company grumbled for quite some time that if you tie an EVA system to

compensation, you could drive the company to do the wrong thing for the

policyholder. That has to do with the integrity of the company. If you weren't

doing the right thing in the first place, this system will not make you do the right

thing or the wrong thing. But it's clear that if we made the decision to lower

profit margins, the EVA would come down in our system.

Mr. Jermyn: I have two more questions. Who exactly determines when there is

a change in assumption? Is it you, the chief actuary? Is it someone else?

Clearly, that's going to have tremendous implications in terms of incentive

compensation.

Mr. de Palo: A lot of this has to do with how you tie EVA into compensation.

The way we're setting up the incentive compensation desensitizes that to some

extent because it's based on the long-term trend. We've created these units that

pay out over a long period of time, so that a one-year change up or down doesn't

affect it. We didn't want to create a system that paid out at a single point in

time. That way, a CEO nearing retirement could drive the assumptions to

increase his value to get a payout just before he retires. The way this system is

being set up for compensation, it's actually spread over ten years. So if you tried

to force up the assumption to increase EVA, but it proved not to be true, it

removes a lot of the incentive.

The actuary is looking at yearly changes in gains or losses, recommending the

change in assumptions, and explaining to management why it's good. It's going

to be a very visible event when this happens, so the actuary better have his or

her facts when making that change.

Mr. Jermyn: What I'm hearing is that, at least for The Guardian, this is for long-

term incentive compensation as opposed to current-year incentive

compensation.


Mr. de Palo: That is correct.

Mr. Brown: In terms of assumptions, when you first set up an embedded value,

you have to go to set all these assumptions and then do it the next year and the

next year. What evolves over a period of time is the actuary defines a

methodology to set assumptions. And once you define a methodology, then

different actuaries are more likely to come up with the same set of assumptions.

Then the changes in assumptions reflect real changes in the real underlying

world, as opposed to actuaries interpreting the same data differently.

Mr. Jermyn: My final question is practical. To what levels within the company

do you think it makes sense to apply value added? Particularly if you look at the

distribution channel, do companies apply it to distribution channel A versus B or

to a career agent channel versus a general agency?

Mr. de Palo: In The Guardian, we do it by profit center. I want to add one other

comment on it as practical experience. We have a new line of business. We

entered a group 401(k) line of business. This will show where statutory, GAAP,

and value added differ so much. On a statutory basis, because it's only a three-

year-old line of business, the 401(k) is losing a substantial amount of money from

new business strain and not yet at critical mass. On a GAAP basis, it's losing a

lesser amount of money. On a value-added basis, all the new business is

profitable and we've covered initial expenses much earlier than either a statutory

or GAAP basis would show.

If you're starting new lines of business, I think EVA methods give management

much more assurance that they're on target with the viability of that line. If

you're not on target, it's also going to tell you that bad things are happening and

that maybe you're never going to get to critical mass. But it gives you useful

information.

From the Floor: This is a very interesting presentation. Armand, I had one

question on the modeling part of this. A lot of companies, and certainly our

company, have built all sorts of models for ALM work. Is that the basic platform

that you work from? Or are these completely separate models that we have to

develop?

Mr. de Palo: We're using the tools we used for ALM. But, for EVA, we're

making different decisions. Our reinsurance profit center uses the TAS. They're

using the same model as for cash-flow testing. The life model is actually using

the same tools, but it is valuing a smaller subset of policies. Our cash-flow

testing uses a very large model. At this point in time, we haven't built this model


the same size. I think, as time goes on, these models will become as robust as

the cash flow models. I think, if you ever read Jim Reiskytl's Financial Dynamic

Surplus Management, you can be glad you have this model. It will let you

produce the future statutory view that Jim has been talking about for the last 10

years.

Mr. Mark A. Milton: We implemented value added about three years ago on a

formal basis. The primary objective is for management compensation. I have

one question related to the change in assumption. One assumption that seems

to be annoying is the interest yield that you actually earn. Our current process is

to leave the discount rate set at what we see as cost of capital, and we have not

changed that through the years. We set our interest-rate on January 1 of each

year based on the current yield curve in effect at that time and then grade over a

three-year period to a long-term historic set of yields. And each year a variance

has to be done. I used the word "annoying" a while ago. It's just something that

has to be done. I'm just wondering if anyone else is using another approach. Is

there a way out of it? Thinking about where yields may be on January 1 this

year, it could have a very significant effect.

Mr. Brown: The most important thing is to make sure that your discount rate

and your interest yields are consistent. You can view your discount rate as being

a risk-free rate plus an element of risk in the business. And your yields are risk-

free rates plus an element for risk in the bond market. As long as those two

move consistently, you tend to get offsetting changes, so the impact on your

embedded value is not material-or at least not as big as it would be if you

moved the two independently.

Mr. Erhardt: In our model, where we have products that have an agreed-upon

spread, we build the model around the agreed-upon spread and report the

variance from spread each year. But the discount rate we do plan to change

periodically. We may not change it each year, but if we get into a low interest-

rate environment, as some people fear, we will probably take this rate down.

That will actually have the effect of increasing EVA for many products. The

scarier scenario is interest-rates spiking up and having to have a higher discount

rate. And when you have a higher discount rate, your expenses basically don't

change, but the present value of future earnings will become a lot smaller item

unless you increase spread. In high interest-rate environments, you may find that

product lines that were profitable in low interest-environments just are not

profitable in turbulent times.

Mr. Milton: I had a couple of other general comments and/or questions. I did

appreciate the panel's discussion. I think a lot of the issues were things that we


had run into and are still struggling with a little bit. One question is, how many

companies do you see taking this type of approach? How many are taking the

EVA approach that we, as actuaries, use versus the EVA approach from Stern

Stewart? And how do you think they're really interpreting the results? Armand,

you mentioned that you looked at trends in embedded values as opposed to an

embedded value of minus the cost of capital, which would be showing negative

numbers. I suspect most companies would have negative numbers if they

looked at that. That may be difficult to get across.

You talked about value added being perceived as the actuaries' "black box."

The actuaries liked that, but no one else did. But going through the actual

verification of results with statutory accounting, our accountants are starting to

buy into it at this point. And we've had some interesting results. When you

have subsidiaries that have been beaten down based on statutory or GAAP

earnings, but appear to be providing a better value-added return, that's opened

some eyes. One thing that was indicated, also, was that, if you hire a large

agency, and because of your compensation structure, that particular agency's

business doesn't earn your hurdle rate. I'm not sure that I'd have the courage to

tell management that 20% of our production is coming from an unprofitable

source. Maybe that's what consultants are for.

Mr. de Palo: We haven't, yet, taken it agency by agency. We use the overall

average agency cost that the company incurred in a year. But we're still

wrestling with the question of how to value noninsurance subsidiaries. We

haven't come to a final answer on that.

Mr. Brown: I think there are only a few companies using the Stern Stewart

approach, although some are thinking about doing it because it's relatively

straightforward and painless. If you view GAAP as being the primary measure,

then why not do the Stern Stewart? It's giving you additional information. In

terms of the actuarial approach, I think the majority of companies that are doing

it in the U.S, or a large proportion, are those that have overseas parents and have

had it forced upon them. But there is a trend. More companies are doing it for

business reasons as opposed to having it forced upon them. The number of

companies making the decision to do it on their own is not significant at the

moment in the U.S.

In terms of understanding embedded value, it's the trend analysis that's

important. The change in numbers from period to period doesn't tell you a great

deal. It's understanding what's driving the change in value that is useful

information to management.


Mr. Erhardt: From all the companies that we're speaking to, it's not in

widespread use at this time. The majority of companies make their decisions on

GAAP or are moving toward GAAP away from statutory. And one of the biggest

advantages I've seen is that there are obviously companies maintaining high

dividends, exceeding what they can pay out because they're not earning their

dividend rate. If managers were basing their considerations on an economic

basis, they might come to a different conclusion on what they're doing with that

business. So it comes back to having good economic data coming to senior

managers so they can make the right conclusion about their business. And that

information isn't coming to management on a widespread basis.

From the Floor: I'd like to follow up on Armand's comments regarding changing

discount rates in response to changes in the interest-rates. For companies that

sell a lot of, let's say, fixed annuities or universal life, how would you address

the shortcoming of EVA in terms of capturing the impact of disintermediation on

projected excess lapses, as well as the impact of spread deficiency of risk, in

terms of the assumed long-term profit spread margins? Perhaps some companies

may have already thought of incorporating, for example, David Becker's option

adjusted value of distributable earnings by running it on a bunch of stochastic

scenarios.

The second question I have is probably food for thought. There has been a lot of

discussion regarding marking both assets and liabilities to market. With respect

to fair value reporting, I get the sense from the people involved in that project-

particularly the FASB-that they seem to be moving toward the direct method,

which is marking assets and liabilities to market separately. Obviously, that's

more like a cash-flow measure of value, as opposed to distributable earnings,

which is an indirect method. And I think there are different methodologies.

Some like the direct method. Some companies would exclude or ignore taxes in

terms of the evaluation of liability cash flows. I seem to have a bias for indirect,

which is distributable earnings. How would you evaluate the impact? When fair

value reporting occurs, I think it's going to change the way companies look at

hedging, risk management, and compensation.

Mr. de Palo: This is not a valuation system. It is a static system to say the least.

It doesn't work on scenarios. It works on statutory accounting. The reason you

don't have to worry about market value to book value is that the assets all run off

the book in the model. It's a long-term project, so short-term changes don't have

much effect. I agree with you that, in a particular year, if there's a run on the

bank or an experience change, it can create very big changes in the value of the

company. But it's not trying to tie into stochastic modeling. And your point is

well taken. Many universal life products on a static model could look profitable,


but on a scenario could look unprofitable. I think what's really happening here

is the actuary has to choose his assumptions well. And if he's changing his

interest assumption and thinks it will create changes in lapses, he probably

should be changing more than his interest assumption in time.

Mr. Brown: What you're trying to measure is economic value. And if there are

risks built into a certain product, that means that you can only capture economic

value by running the model stochastically using whatever approach is needed. If

a certain product has small risks that you can only capture by stochastic

modeling, then it's probably not worth the effort. But if it's a product where the

risk is significant and can only be captured by doing stochastic models, and that

product line is significant in terms of the overall company, then you may need to

think about modeling it stochastically, introducing excess lapse formulas, and

whatever else is needed.

Mr. Hubert B. Mueller: I'd like to add a little bit of a European perspective on

the topic of value added. In Europe, most large companies, and especially all

the multinationals, are using value added very extensively. We have even used

it in Germany and Italy with noninsurance businesses such as determining the

economic value added of bank business and mutual fund business. And it has

been very well-received also by the investment banking community in terms of

determining the true economic value of a company when the transaction takes

place, whether it's for internal or external purposes. So I would like to add that

point, especially in response to Mark's question about whether it's used in

noninsurance businesses. At least in Europe, we are doing that.

Documents

Using Value-Added Information in Practice€¦ · Using Value-Added Information in Practice 5 making value-added decisions. And they had very little synergy occurring among all the